You are viewing the SNL 2018 Archive Website. For the latest information, see the Current Website.

Poster B2, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

Lexical selection and the elusive role of the left IFG: an fMRI study

Hanna Gauvin1,2, Katie McMahon3,4, Greig de Zubicaray1;1Faculty of Health and Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, QLD, Australia, 2School of Psychology and Counselling, Queensland University of Technology, Brisbane, QLD, Australia, 3School of Clinical Sciences, Queensland University of Technology, Brisbane, QLD, Australia, 4Herston Imaging Research Facility, Royal Brisbane & Womens Hospital, Brisbane, QLD, Australia

According to prominent neurobiological models of lexical selection, the left inferior frontal gyrus (LIFG) plays an important role in resolving semantic interference in speech production, performing domain-general, top-down control processes (e.g., Belke & Stielow, 2013; Oppenheim et al., 2010; Schnur et al., 2009). However, a role for the LIFG is not consistently reported across studies using various semantic interference paradigms. Continuous naming studies (patient studies and neuroimaging) have consistently failed to observe significant LIFG involvement. Blocked cyclic naming paradigm studies (neuroimaging, neurostimulation and lesion patients) show equivocal results, with as many studies demonstrating LIFG involvement as ones that don't. Results from the Picture-Word Interference (PWI) paradigm are likewise inconsistent. In this study, we investigated the role of the LIFG in semantic interference in PWI. In particular, we investigated whether response set membership (in which distractor words are also target picture names) influences LIFG involvement. It has been proposed that distractors that are also eligible responses are allocated greater attention, or that non-response set distractors require greater inhibition (e.g. Roelofs, 1992, 2003; Lamers et al., 2010). We therefore hypothesised that LIFG involvement would vary according to response set membership. [Methods] Twenty healthy participants named 40 target pictures while ignoring a superimposed distractor word during fMRI. Each target picture was presented with 4 distractor words that varied according to response set membership (2) x semantic relatedness (2). fMRI data were acquired with a sparse sampling design using a 3T Siemens Magnetom Prisma using a 64-channel head coil. Data were analysed using SPM 12. The following ROI's were selected a priori from the Hammers et al (2003) probabilistic atlas: left pSTG, left MTG, left + right ACC, left PTL and left IFG. [Results] LME analysis of the behavioral data (random: Participants + Items, fixed: Distractor Condition + Response Set Membership) show no significant effect of response set membership. There was a main effect of distractor condition, with the typical semantic interference effect (RTs semantically related distractors > unrelated). No significant interaction was observed. The fMRI data largely replicated the behavioural data: There was no significant main effect of response set membership or interaction with semantic context in the LIFG or other ROIs. In the typical Response Set member condition, significant activation was observed for the Related>Unrelated contrast in the left pSTG and MTG, replicating previous studies. We next combined the data from the two response set conditions to contrast Related> Unrelated conditions, this time revealing significant activation in the LIFG. [Conclusion] Our findings suggest that LIFG involvement does not differ markedly between response set and non-response set distractor manipulations. That significant LIFG activation was observed only when data from response-set membership conditions was combined suggests the inconsistent results reported in the literature might reflect signal-to-noise or statistical power issues. The findings partly support models of speech production that assume strengthening of conceptual-lexical connections and interference effects when distractors are response set members compared to non-response set members.

Topic Area: Control, Selection, and Executive Processes